We have been studying methods of testing for genetic association with disease. One project is to extend current methods based on the conditional likelihood to allow restrictions on the parameter space. In genetic association studies, it is often thought that the relative risks of one versus two copies of a risk allele are not completely free to vary. Different restrictions could reasonably be placed on the parameter space, resulting in more powerful tests. A paper has been published. Ongoing research is focussed on methods of multiple comparison adjustment for triad association studies.